Competing with Giants: Colin Earl, CEO of Agiloft (Part 3)

Colin Earl: The next thing that happened was somewhat of a surprise. It was being picked up by very large businesses. I won’t mention by name but it’d suffice to say that one client was one of the world’s largest defense contractors. When a guy representing them called, I asked him what they will be using it for.

Then he replied, “We’re going to be using it for satellite communication errors.” Then I wondered, “Why? You’re a Fortune 50 company. You must have something for tracking satellite communication errors. Satellite is one of your primary areas of business.” The voice got really cold at the end of the line. He said, “Yes. If I want to add one field, it takes two months for IT to do it. I have to call in political favors to get it done at that time. Now, I downloaded your product this morning, and I added all the fields I needed this afternoon. Would you take a credit card?”

I realized then that we’d built Support Wizard to be deeply configurable. That was just my inclination as a software guy – to make things flexible. We didn’t honestly have a clear business rationale as to why this was important. Indeed, it was very different from the norm at that time. From my background, it seemed like the right thing to do. As a result, we were making sales to large complex organizations where the business people or the IT people were frustrated with the lack of flexibility that they’re getting from the established vendors.

After some time, we took a step back and said, “Maybe this is the market we should be going after. What could we do to build a product which allowed IT managers and IT staff to build and configure complete enterprise systems? Not just customize support product for their particular needs, but build entire new applications without coding.” If this guy had to code, he wouldn’t have been able to put in the fields he needed in an afternoon.

In general, anytime you introduce coding, you introduce custom bugs. You introduce communication gaps between the developers who are producing the result and the business people who are requesting it. We stepped back and designed, from scratch, a platform that would allow design configuration of large enterprise systems without coding. This was an extremely ambitious project. We were really betting on Moore’s Law.

Again, this is a worthwhile insight that if you’re aiming to address a large market or something that makes a dent in the world, it’s going to be a significant effort. If it was easy, somebody would already have done it. That type of effort is likely to be a multi-year effort. You can’t build it or optimize the design for technology as it exists today. You have to look at where technology is going and base your design on where you think it’s going to be when your product is ready for market.

In our case, we built on the expectation that Moore’s Law would deliver computers with multiple CPU cores with 8GB or RAM or more at a reasonable price point. At the time we started, this really was not true. At that time, Oracle was running its primary server to serve the entire organization on 2GB of RAM. Those 2GB of RAM cost hundreds of thousands of dollars. We were really building for the future and, of course, Moore’s Law delivered as it tends to. The net result was that we were able to design a product which was extremely flexible, fast, and scalable.

The tradeoff was that, to run efficiently, it needed at least 8GB of RAM and four CPU cores, which five years ago, was an absurd requirement. By the time we finished, it was basically commodity hardware. Since then, we’ve grown that platform. Reflecting on the flexibility that we’ve built the agility that we were aiming at, we built our first product to address the contract management market in the space of four weeks from scratch on the platform. That product won PC magazine’s Editor’s Choice Award as the best Contract Management system on the market. This is a multi-billion dollar market.

This segment is part 3 in the series : Competing with Giants: Colin Earl, CEO of Agiloft
1 2 3 4 5 6